Auto Scaling Group

Auto Scaling Group contains a collection of EC2 instances that are treated as a logical grouping for the purposes of automatic scaling and management.

Auto Scaling Group uses Auto Scaling features such as health check replacements and scaling policies.

The purpose of Auto Scaling Group #

• Scale out (add EC2 instances) to match an increased load
• Scale in (remove EC2 instances) to match a decreased load
• Ensure we have a minimum and maximum number of instances running
• Automatically register new instances to the Load Balancer
• Replace unhealthy instances
• Cost Saving (only run at optimal capacity)

Creating an Auto Scaling Group #

• Create Launch Template
• Create Auto Scaling Group
  • Select Availability Zones
  • Select Availability Zone distribution
  • Attach to an existing Load Balancer
  • Turn on Elastic Load Balancing health checks
  • Define the desired capacity
  • Set up Automatic Scaling (optional)
  • Select Instance maintenance policy
  • Additional capacity settings
  • Additional settings

Auto Scaling Groups - Strategies #

• Manual Scaling - update the size of an ASG manually
• Dynamic Scaling - respond to changing demand
  • Simple / Step Scaling
    • When a CloudWatch alarm is triggered (i.e. CPU > 70%) then add 2 instances
    • When a CloudWatch alarm is triggered (i.e. CPU < 30%) then remove 1 instance
  • Target Tracking Scaling
    • Example: Average ASG CPU to stay around 40%
  • Scheduled Scaling
    • Anticipate a scaling based on known usage patterns
      • Example: increase the min. capacity to 10 at 5pm on Fridays
  • Predictive Scaling
    • Uses Machine Learning to predict the load

» Sources « #

• https://docs.aws.amazon.com/autoscaling/ec2/userguide/auto-scaling-groups.html

Full YouTube Rahul’s AWS Course: https://www.youtube.com/playlist?list=PL7iMyoQPMtAN4xl6oWzafqJebfay7K8KP

» References « #

• Scalability & High Availability
• Elastic Load Balancing

» Disclaimer « #